Figure 1.
Map of the study area with the location of three large (MB, CH, NG) and two small (FU, ND) indigenous forest fragments that hold breeding populations of cabanis’s greenbul (Phyllastrephus cabanisi).
Analyses were performed along linear transects at two different geographical scales: a local (within-fragment) scale in fragment MB (A), and a landscape (among-fragment) scale including fragments CH, NG, FU and ND (B).
Figure 2.
Distribution of maximum post-fledging recapture distances of P. cabanisi fledglings during their first year and over multiple years.
Data were collected between 2007–2012, sample sizes are indicated above each vertical bar.
Figure 3.
Levels of admixture provided by Bayesian clustering analyses (STRUCTURE) for each period separately (left panel: period 1/right panel: period 2).
Following the procedure outlined in [56], optimal cluster sizes of K=2 and K=4 were assigned to period 1 and period 2, respectively.
Figure 4.
Multiple distance class plots illustrating the gradual decay of positive genetic structure with increasing geographical distance.
Plots are shown for each sex at various spatio-temporal scales: at a regional scale for period 1 (a), period 2 (b) and both sexes lumped (c); at a local scale (within Mbololo) for period 2 (d). Error bars indicate 95% CI as determined by 9999 bootstraps.
Figure 5.
Distribution of randomized statistics under the null hypothesis of equal dispersal rates in both sexes.
Permuted test statistics are ‘male mAIc-female mAIc’ (left) and ‘vAIc female/vAIc male’ (right). Observed statistics are indicated by vertical dashed line and extreme positive values indicate female-biased dispersal. Upper panel: regional - period 1; middle panel: regional - period 2; lower panel: local - period 2.